PSY236 Lecture Notes - Lecture 3: Latent Inhibition, Operant Conditioning, United States Naval Observatory
Week 3 PSY236 Lectures:
Classical Conditioning 5 - Contingency
Contiguity or contingency?
• Contiguity = the connectedness in time and space of the CS and US
• As the delay increases between the CS-US so the rate of learning decreases
• But the CS also becomes less useful as a predictor of the US
• Contingency = the predictability of the occurrence of one stimulus from the presence of another
o Contingency = probabilistic relationship with US given that a CS has occurred refers to the
extent to which the pairing of the CS with the US is necessary and sufficient for learning to
occur
So contiguity is necessary…
• But NOT sufficient for classical conditioning to occur
• There must also be a consistent relationship or correlation between the CS and the US
• To experience a reliable correlation between the CS and the US the subjects must be exposed to
numerous instances of the CS and US, thus many trials are typically necessary for conditioning
• Most learning curves support this, especially in appetitive learning situations
• Aversive conditioning – they do not wait to learn the relationship – once was enough
So contiguity and contingency is necessary –
• Kamin’s seminal study on blocking showed that contiguity alone was not sufficient
• Blocking
o Phase 1: train light CS
o Phase 2: train light CS in compound with neutral (but equally salient) metronome CS
o Phase 3: test elements of compound individually: pre-trained light blocks conditioning to
metronome
• Kamin’s procedure and results
o Pre-training Noise blocked light from → CR
o NO pre-training of noise → no blocking, so L → CR
Why didn’t light → a CR?
• It had been paired (in compound with Noise) contiguously with the US?
• Because Noise already predicted that the US would follow
• Learning during a conditioning trial is a function of predictive error: the discrepancy between the
actual outcome of a conditioning trial and the expected outcome of that trial
• Unblocking occurs because there is a discrepancy between expectation and reality
Contingency will depend on:
• Reliability of CS-US pairing:
o How often is the CS followed by US?
o What is the probability that the US will occur given that the CS has just occurred
• Uniqueness of CS-US pairing:
o How often does US happen without CS?
o What is the probability of the US occurring given that No CS has occurred?
Contingency
• Refers to the predictive relationship between stimuli (CSs and USs)
• The CS has to convey information about US occurrence. ‘Unpacking’ the terminology of Rescorla to
describe these relationships:
o The probability (p) of a US occurring given that (/) a CS is present, is greater than the
probability (p) of a US given that NO CS is present
• Rescorla’s equation describing a positive correlation between the CS and US
find more resources at oneclass.com
find more resources at oneclass.com
• p(US/CS) > p(US/No CS)
o The left side of the equation notes the percentage of CSs that are temporally contiguous
(paired) with a US
▪ If p = 1.0 then 100% of CSs are paired with USs
▪ If p = 0.5 then 50% CSs are paired with UCSs and 50% of CSs are presented alone
▪ If p = 0 then all CSs are presented alone, there are no CS-US pairings
o The right side of the equation notes the percentage of time intervals without a CS in which a
US occurs
▪ If p = 1.0 then USs are presented on 100% of time intervals with No CS present
▪ If p = 0.5 then USs are presented on 50% of time intervals with NO CS present
▪ If p = 0 then USs are never presented when NO CS is present
No contingency between CS and US
• If the CS is an unreliable predictor of the US, then the CS and US are not correlated
Negative contingency between CS and US
• If the CS reliably predicts the absence of the US, then the CS AND US are negatively correlated
The effect of contingency on classical conditioning:
• Group A vs. Group B
• Bell rings, shock occurs and no shock can occur → group B = conditioned fear of the bell
• For both groups, there’s only 40% chance that bells will be followed by shock. However, for group B,
shock is less likely when no bell is sounded and for this group, the bell becomes a fearful stimulus
Positive contingency
• P(US/CS) > p(US/NO CS)
• 40% > 20%
• Bell → Fear CR
Negative contingency
• P(US/CS) < p(US/NO CS)
• 40% < 80%
• bell → inhibitory stimulus
• shock less likely to occur after CS
So how do we calculate the relative probability of learning occurring?
• First take notice that timeline is divided into 12 equal intervals of time
• Next we calculate the left side of the equation
• There are 4 time intervals with a cs
• A US occurs in all 4 CS intervals
• Therefore the probability of a US given the presence of a CS is 1.0
• Next we wil calculate the right side of equation
o There are 8 time intervals with NO CS
o A US occurs in 0 of these no-CS intervals (i.e. 0/8)
o There probability of a US given the absence of a CS is 0
• This is called EXCITATORY CONDITIONING
Inhibitory conditioning
• When the relationship is negative (<) instead of positive (>)
Summary:
• When subjects experience CSs and UCSs that are positively correlated they acquire a conditioned
response to the CS; this is called excitatory conditioning
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Contingency: refers to the predictive relationship between stimuli (css and uss, the cs has to convey information about us occurrence. If p = 1. 0 then 100% of css are paired with uss. If p = 0. 5 then 50% css are paired with ucss and 50% of css are presented alone. If p = 0 then all css are presented alone, there are no cs-us pairings: the right side of the equation notes the percentage of time intervals without a cs in which a. If p = 1. 0 then uss are presented on 100% of time intervals with no cs present. If p = 0. 5 then uss are presented on 50% of time intervals with no cs present. If p = 0 then uss are never presented when no cs is present. If the cs is an unreliable predictor of the us, then the cs and us are not correlated.